The aim of this study was to determine whether serum fructosamine is associated with the risk of type 2 diabetes mellitus (T2DM) in a cohort of middle-age Finnish men. Fructosamine and other cardiometabolic risk factors were measured in a sample of 2,520 subjects from the population-based Kuopio Ischemic Heart Disease study. Associations between baseline fructosamine levels and risk of T2DM involved Cox-regression analyses, progressively adjusted for potential confounders and mediators. Mean baseline age and fructosamine levels were 53 years and 267 µmol/L, respectively. During a median follow-up of 23 years (interquartile range 18-25 years), 153 (6.1 %) new cases of T2DM were recorded. We observed an approximately "J-shaped" association between fructosamine values and T2DM risk, with a nadir at about 250 µmol/L. In subjects with fructosamine ≥250 µmol/L, the hazard ratios per 1 standard deviation (28 µmol/L) higher fructosamine levels resulted 1.20 (95 % CI 1.00-1.44; p = 0.046) in a model adjusted for age, body mass index, systolic blood pressure, serum high-density lipoprotein-cholesterol and family history T2DM. This association did not materially change upon further adjustment for C-reactive protein, physical activity, serum triglyceride and insulin resistance. However, no association was found when adjusting for baseline fasting plasma glucose (FPG). In this cohort of middle-age Finnish men, we found a "J-shaped" relationship between fructosamine and risk of T2DM, with a higher risk in subjects with baseline levels of fructosamine ≥250 µmol/L. Such an increased risk, however, was not independent from baseline FPG levels.